Research On Performance Enhancement Of Streaming Media Application

As network communication technology and multimedia technology developing, streaming media technology has developed dramatically, and streaming media application has been widely used in many areas. However, traditional Internet is a best effort network, which was designed for traditional data transmission at first, and could not guarantee Quality of Service (QoS) for streaming media application.To support streaming media application with QoS, research has been done in two aspects:One is from physical and network layer, to let Internet support QoS for streaming media, including research on network equipment, network protocol and network topology. The other one is from application layer, using QoS optimizing strategies at streaming server side and user side to enhance network adaptability for streaming media application.The main work and contribution of this paper is that a QoE supported streaming media system model has been proposed. It is based on traditional streaming media system model, but adding new function modules-media filter and feedback process. The new QoE supported streaming media system model can meet the requirement of heterogeneous terminals, which have different screen size, resolution and color attribution, and supply QoE support. For example, when media filter receives feedback information which requires media stream with high QoE, media filter can adjust sending priority to a high value in order to get a high throughput. On the other hand, when media filter receives feedback information which requires media stream with low QoE, media filter can adjust sending priority to a low value in order to use network bandwidth effectively. This paper describes each function of this QoE supported streaming media system model in detail, and designs the algorithm for core module. A test bed and test cases are used to evaluate the performance of this algorithm.